Neural Direct Adaptive Active Disturbance Rejection Controller for Electro-hydraulic Servo System

This paper develops a neural direct adaptive active disturbance rejection controller for the electro-hydraulic servo system (EHSS). EHSS in the construction mining machinery is a high-order nonlinear system with uncertainties and heavy external disturbance, which challenges the controller. The proposed control scheme, integrating neural direct adaptive controller and linear extended state observer (LESO), is designed based EHSS's reduced-order model. The stronger robustness and the improved tracking performance can be expected by using LESO to compensate for remaining uncertainties. In addition, the rationality of reduced-order model and the stability of the proposed controller are proved. Comparative simulation results show that the controller has an excellent position tracking performance.